1. Support to Anti - Corruption in Sierra Leone. Progress Assessment Report November 2018

17. 16 Respondents in Urban Areas were slightly less likely to have paid a bribe for Education services durin g the last year. Whether the respondents lived in an area with intensive outreach or not did not influence the likelihood of whether they paid a bribe or not. Have you paid a bribe within the last year for Police services ? Slightly more respondents report paying a bribe to the Police (52%) within the last year than the other sectors. Hence, the frequency of bribe paying hardly differ s , despite the public perception is much more positive for the other sectors. This may be due to the fact that relatively few people has their own vehicles or bikes – which means that they experience bribe taking, but they are personally not the bribe givers. C hart 31 : Have you paid a bribe within the last year for Education Services : C hart 32 : Have you paid a bribe within the last year for Education Services and Community type : C hart 33 : Have you paid a bribe within the last year for Education Services and PNB outreach level :

6. 5 PWD were also interviewed, although not proportionately. 17 % of respondents were recorded as having a disability (compared to 16% in previous assessment) , a figure significantly higher than the official statistic of 1.3% for Sierra Leone 4 . This anomaly was a result of CSO Animators being asked to include persons wit h disability in the data collection. Western Area Rural recorded a higher percentage of PWD; 32 % of respondents having a disability. Little demographic variance was noted between districts; neither in “type of Community”, gender or age. 4 “ Sierra Leone 2015 Population and Housing Census - Thematic Report on DISABILITY” https://sierraleone.unfpa.org/sites/default/files/pub - pdf/Disability%20Report.pdf page 11) Chart 5: Respondents with disabilities Chart 6 : Respondents with disabilities per district C hart 3: Responses by gender C hart 4: Responses by age group

18. 17 Neither community type nor level of sensitisation shows any impact on the responses. Perception of the likelihood of ACC acting against corruption 82 % of respondents perceive the ACC to be more likely to act against corruption compared to one year ago ; a similar result to the previous assessment s ( 80% in May, 85% in August) . R elatively high and stable level s suggest that the ongoing priority given to anti - corruption by the new government plus extensive PNB outreach and communications work has contributed to creat ing a stronger ACC public profile and greate r public awareness . C hart 34 : Have you paid a bribe within the last year for Police Services : C hart 35 : Have you paid a bribe within the last year for Police Services and Community type : C hart 36 : Have you paid a bribe within the last year for Police Services and PNB outreach level :

10. 9 Sierra Leone Police R espondents’ perception of Police officials continues to be lower than those of Health and Education officials. The third assessment, however, continue s to indicate positive trends; the ce ssation of bribe taking by police increased from 6% in May , to 13% in August , and to 18% in November. The percentage of respondent s not experienc ing a change fell from 45% to 38%. In line with health and education, respon dents from PNB Core areas perceive more positive change to the behaviour of police officials than those from Hub and Spoke and Radio Only Area s w ith a smaller percentage of respondents see ing no change at all. C hart 16 : Perceived behaviour change by police officials in PNB Core areas C hart 17 : Perceived behaviour change by police officials in Hub and Spoke areas Chart 15 : P erceived behaviour change by police officials

16. 15 Frequency of paying bribes Have you paid a bribe within the last year for Health services ? 45% of respondent s report hav ing paid a bribe for health services within the last year. This was particularly noticeable in Bombali and Kono District, while less in Bo and Kenema. Respondents from rural areas report bribe payment more frequently than respondents from urban/mixed localities. The outreach level d oes not appear to influence . Have you paid a bribe within the last year for Education services ? Despite significant improvement to public perception of Education official’s behaviour, 39% of the respondents report hav ing paid a bribe for Education services within the last year. This appears to contradict the data on perception of behaviour change, which was very positive. In this res pect, it should be noted that the period in question includes the period prior to the Free Quality Education initiative . C hart 28 : Have you paid a bribe within the last year for Health Services : C hart 29 : Have you paid a bribe within the last year for Health Services and Community type : C hart 30 : Have you paid a bribe within the last year for Health Services and PNB outreach level :

5. 4 Demographics The questionnaire was administered in both urban, rural and peri - urban communities (less den se populations located at the fringes of urban settlements with productive land). There are similar response distribution s to previous assessments , with one notable variation: an increase in samples from Urban communities (42% to 48%) and a corresponding decrease in samples from Rural (37% to 29%). T he data collected in th is t hird assessment thus corresponds more closely to the Sierra Leone demographic 3 . Disaggregation by type of locality per District remains similar to the previous assessment. Unsurprisingly, 95 % of responses In Wester n Area Urban were urban. Bombali, Kenema and Kono cover a greater mix of communities, with Kono having a higher rural response distributi on . The CSO enumerators were asked to aim for a 50/50 gender distribution plus equal age distributions. Gender balance was almost achieved across age groups; age distribu tion closely reflect s PNB data collated on the reporting platform. Both variables are similar to the previous assessment. 3 The 2015 Population and Housing Census showed that 41% of the Sierra Leonean population lives in Urban area (page 30) https://www.statistics.sl/images/StatisticsSL/Documents/final - results_ - 2015_popul ation_and_housing_census.pdf Chart 1: Responses in types of Community Chart 2: Responses in types of Community by District

9. 8 As with the health trend, respondents in PNB Core areas s ee more positive change to education provider behaviour than in Hub and Spoke and Radio Only Areas. In Core areas 52 % say that education providers ha ve stopped requesting bribes, 31% in Hub and Spoke and 50 % in Radio only areas. That Radio only performs better than the Hub and Spoke areas, can be caused by the Radio only areas having a relative small sample size. Western Area Urban recorded the most positive response to a change in behav iour by education providers. Here , 72 % of respondents have indicated a c essation in payment requests and just 4% no change. This is a significant shift from the first progress assessment conducted in May 2018 w h ere in Western Area 44% of respondents had stated no change in behaviour . This indicates that the implementation and impact of the Free Quality Education has been more successful in the Western Area Urban compared to other part s of the country. This reinforces the experience of the call centr e, which reports challenges in the provinces with parent s not being able to distinguish between Government approved and Community Schools, where parent have to pay fees for the latter. Chart 11: P erceived behaviour change by education officials C hart 12 : Perceived behaviour change by education officials in PNB Core areas C hart 13 : Perceived behaviour change by education officials in Hub and Spoke areas C hart 14 : Perceived behaviour change by education officials in Radio only areas

19. 18 In both Kenema and Kono however significantly fewer respondents consider it likely that the ACC would act against corruption , compared to August 2018. Both Bo and Bombali continue to maintain high level s of public confidence. Respondent s are more positive in PNB Core Areas on the likelihood of ACC taking action against corruption ; with little variation between PNB Core and Radio Only . Respondent demography does not suggest significant difference to the perceived likelihood of the ACC acting against corruption. Table 12 : Demographics and perceived likelihood of ACC acting against corruption Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes No 18% 18% 18% 19% 17% 18% 18% 15% Yes 82% 82% 82% 81% 83% 82% 81% 85% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% C hart 37 : Perceived likelihood of ACC acting against corruption C hart 38 : Perceived likelihood of ACC acting against corruption by outreach level

13. 12 The conclusion of the May and August assessment report s that “ MDA engagement in the PNB Campaign is more effective in the Provinces “ remains valid for the Health and Police sectors . A strong contributory factor will be PNB engage ment with district s ervice providers through the monthly Regional Technical Working Group s and Accountability Forums . These are strengthen ing transparency and accountabilit y processes close to the point of service delivery. In contrast, Western Area engagement with MDA HQ s has placed greater emphasis on programmatic matters in relation to the implementation of the PNB, and less on the issue identified from the PNB data. This is partly due to the MDA with greater distance from the service delivery . In the Education sector , the introduction of Free Quality Education has redef ined the relationship between service provider and the public . As such, it appear to contribute more significantly to changing public perception s than localised MDA action s. Since the implementation of FQE is more comprehensive in Freetown and urban centres than in rural areas (where most schools are Community Schools), the perceived positive behaviour change is much greater in Freetown and other urban centres . C hart 21 : Perceived behaviour change by educations officials in the Provinces C hart 22 : Perceived behaviour change by education officials in Western Area C hart 23 : Perceived behaviour change by police officials in the Provinces C hart 24 : Perceived behaviour change by police officials in Western Area

14. 13 Resistan ce to bribery The data suggests that r espondent s are more likely to say NO to pay ing a bribe across all 3 service sectors. In May 60 - 7 0 % stated they were more likely to resist, increasing to 70 - 80% in August and in 80 - 90% in November. Table 8 : Are you more likely to say NO to pay a bribe now than one year ago? Health Education Police No 219 11% 158 8% 343 17% Yes 1797 88% 1856 91% 1670 82% (blank) 24 1% 26 1% 27 1% Grand Total 2040 100% 2040 100% 2040 100% In areas with intensive outreach, the respondents generally are more likely to resist paying a bribe, however by a s lim margin. As in August ’s assessment, the level of PNB outreach appears to ha ve limited effect on whether people are likely to resist bribery or not. However, it seems, as illustrated for Health in Chart 8 - 10 and for Police in Chart 16 - 18, that the PNB outreach has a greater influence on service provider behaviour . Hence, the data suggest that the outreach has more impact on the Service Providers behaviour than the citizen’s likelihood of refusing to pay a bribe when requested. C hart 26 : Change in Likelihood of saying NO to bribes for Education C hart 25 : Change in Likelihood of saying NO to bribes for Health Service C hart 27 : Change in Likelihood of saying NO to bribes to Police :

24. 23 MDA Action s Across the MDA s , more respondents are able to mention an action taken by them to stop bribery : w ithin the Education sector this increased from 55% to 66% ; i n the heath sector from 35% to 46% mention an action taken by the MDA s , f or the SLP it is 29% to 33% of respondents. Similarly , the percentage of respondents unable to mention an action taken by any MDA within the last 3 months , fell. This data is consistent with findings from previous sections in this report on perceived behaviour change among service providers . T he public consistently report the most positive change among education providers, followed by health providers, then Sierra Leone P olice. The data d oes not indicate a clear relation ship between the level of outreach and respondents ’ ability to describe MDA action s within the previous 3 months , although it is worth noting that PNB core areas perform consistently better than the Radio only areas. In general, respondents in Bombali were able to mention the most actions taken , followed by Kenema, w ith the least percentage of respondents able to mention an action in Western Area (Urban and Rural). This is likely caused by the lack of a TGW dealing dir ectly with issues of service delivery in the Western Area . C hart 51 : Number of respondent who could mention an action taken by MDAs C hart 52 : Percentage of respondent who could mention an action taken by MDAs by PNB outreach level C hart 53 : Number of respondent who could mention an action taken per district

7. 6 Perceived behavioural change among Public Service Delivery Officials Respondents were asked whether they had noted behaviour change among service delivery providers in the last 3 months within Health, Education and Police sectors. If they had perceived a change, they were asked to identify what kind of change, and offered a multiple choice listing with the opportunity to select more than one option. Hence, in the following charts perceived changes ( green and red bars) and the No change should be seen as separate figures totalling more than 100%. The percentage calculation of both the perceived change (green and red bars) and the No change is based on the total number of responses made by individual respondents. Health The data suggests positive improvement to public perception of health officials’ behaviour. 8 1 % of respondents have experienced some level of behaviour change among hea lth providers within the last 3 months, a 10% increase since the second assessment . The m ost commonly perceived change s continue to be service providers demand ing bribes less often (38 %) ; with greater secre cy (3 6 %) ; and cessation of bribe demands (30 %). Only 1 % state that that health providers now request higher payments/bribes or with greater frequen cy - an improvement from 4% in the fir st assessment and 2% in the second . Citizens interviewed in c ore PNB chiefdoms – where PNB outreach by ACC/CSOs has been extensive – noted mo re positive behaviour change tha n citizens reached more occasionally through Hub and Spoke or by Radio only . In the core area s 32 % responded that health officials had stopped asking u s for bribes , compared to 16% in Hub and Spoke and 23 % in Radio Only. Chart 7: P er ceived behaviour change by health officials

8. 7 17 % of citizens in core area ha ve seen no change at all, 21 % in Hub an d Spokes areas, 29 % in the area only reached by Radio. Furthermore, core areas contain the highest percentage of people perceiv ing that health official s have stopped asking for bribes ; 32% compared to 16% in Hub and Spoke and 23% in Radio only areas. As in previous reports , data does not suggest significant variance between Rural, P eri - Urban and Urban communities. H ence , reported improvement to Core areas does not relate t o community type. Bombali r esponses show the greatest improvement ; 46 % indicat e that health official s ha ve stopped asking for bribe s , compared to16% in Kenema. In all three progress assessment s Kenema Urba n area has performed significantly less well than others . The Integrity Management C ommittee in Kenema Government Hospital is weak with less robust systems to prevent bribery. For example, c entralised payments are not in place; a standard in most Government Hospitals. Education Respondent perceptions of behaviour change among education providers continue s to improve. I n the first assessment 20% of respondents indicated that Education officials had stopped requesting bribes ; this rose to 40% in the second assessment and 50% in the third. Over the period, t he percentage of respondent s not perceiv ing any change declined f ro m 19% to 8%. Th is positive trend is best explained by the introduction of the free education scheme . This has change d dynamics between education providers and the public. As a result , citizen perception of behaviour change by education officials is greater than that for Hea lth and Police. Chart 8: P erceived behaviour change by health officials in PNB Core areas Chart 9: P erceived behaviour change by health officials in Hub and Spoke areas Chart 10: P erceived behaviour change by health officials in Radio only areas

2. 1 Pay No Bribe Progress Assessment Report: Citizen Perceptions of Behaviour Change by Public Official s in the Health, Education, and Police sectors . November 2018 Summary This third and final report highlights progress with respect to outcomes and implementation. The three 2018 progress Assessments conducted in May, August and November 2018 indicate significant improvement to public perceptions of behaviour change by public officials in the Health, Education and Police sectors. The assessments suggest that this ch ange has been driven by four main factors, the first three of which are mutually reinforcing; • MDA engagement and actions taken through the PNB Campaign, • Increased citizen knowledge through sensitization on rights and obligations • Increased public confidence in the ACC’s willingness and ability to tackle petty corruption, combined with greater awareness that reporting to the ACC is possible and accessible. • Policy change that alters public engagement with the service provider (exemplified by the Free Quality E ducation). This November 2018 Progress Assessment confirms previous conclusions, mainly : • Improved p ublic perception of service provider behaviour • Greater p erc eption of positive behavioural change by police and health sectors in the p rovinces, compared to Western Area. This suggests that more intensive engagement by ACC at district level continues to yield better outcomes tha n national level MDA engagement . • Perce ptions of positive behavioural change in the Education sector is especially high in Western Area . This reflects the more advanced implementation of Free Quality Education and suggests that major policy change can achieve greater impact than local level action. • D emographic variance regarding people’s perception of bribery does not appear to be influen ced by gender, age or disability. • F indings continue to suggest linkage between PNB implementation and public perception , although this is less strong than could be expected. Where citizens have received intensive outreach, knowledge of PNB and s ervice c harters is stronger, a s is awareness of what MDAs should and should not be doing. Citizens are then better able to identify behaviour changes by service providers a s well as actions taken by MDA s in response to reports. They are also more likely to know how to report a bribe. MDAs in areas where citizens and the ACC are more engaged will come under increased pressure to change practices from more engaged citizens: this is more likely to happen in the districts outside of Freetown as the MDAs are more acti vely engaged the provinces. Education however is an exception , due to the Free Quality Education initiative , which appears most successful in the capital.

25. 24 Knowledge of correct fees for service 32% of respondent had no knowledge of the correct fees for specific services , which is an improvement fr om 45% in the August assessment . K nowledge of Education fees and service s are highest and l east for SLP ; consistent with behaviour change perception and knowledge level of MDA actions. The increasing knowledge level can be attributed by a focus on implementing Service Charters at District level through the Regional Technical Working Groups The youth had slightly higher knowledge levels than older citizens – particularly for Education. Table 18 : Demographics and Respondent who knows about fees for services Gender Age Disability Grand Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes I know about fees for education services 59% 59% 59% 61% 59% 53% 58% 63% I know about fees for Health services 42% 43% 41% 43% 42% 39% 41% 41% I know about fees for Police services 28% 29% 28% 29% 28% 29% 27% 28% I do not have knowledge of correct fees for specific services 32% 31% 33% 29% 32% 37% 33% 29% Source of information on fees and services in education The percentage of Citizens claiming knowledge of fees and services in the education sector is suggestive of Service Charters as a source of information ; a significant increase from 22% to 46% (from 9% of total to 24% of total). Display of service charters has been a focu s for Regional Technical Working G roups since the implementation of Free Quality Education, with a positive impact . C hart 54 : Number of respondents with knowledge of fees for services

4. 3 15 animators x 5 districts x (15 Core - PNB area responses + 10 Limited outreach area responses + 5 Only Radio outreach area responses ) = 2250 Responses The actual number of responses, as Table 1 illustrates, was lower than anticipated in most Districts . This is likely due to mobile data issues occurring simultaneous with the assessment, which prevented some enu me rators from downloading the questionnaire and others from submitting them to the database. These issues had greater impact in the Western Area, where the CSO coordinator did not follow up with the enumerators to the same extent. Th e actual number of responses from this third assessment was 2040, compared to 2312 in the second and 2359 in the first . Table 1: Number of Responses collected per District Bo Bombali Kenema Kono Western Area Rural Western Area Urban Grand Total Planned Number of Responses 450 450 450 450 450 2250 Actual number of Responses 423 466 420 429 70 232 2040 T he same sampling methodology as previous assessment s was applied. T he CSO Coordinators instructed each enumerator to carry out the sampling describe below, and supported it with an instruction on WhatsApp : Sampling “ Please carry out 30 surveys in total. 15 from the core PNB location, 10 limited outreach location and 5 from the “Radio only” location. Ask every 5 th person you meet to collect a response from them. Every time you have conducted an interview, make sure to move some distance, so you do not interview people who are part of the same group. Go to different locations within t he area you are carrying out the survey in. Aim toward collecting responses from 15 females and 15 male (50/50) over the 3 days Aim towards collecting 12 responses from the youth, 12 from the middle aged and 6 from the seniors over the 3 days Aim toward collecting at 3 responses from people with disabilities over the 3 days ” During qu estionnaire administration , CSO Coordinator s monitored the process in their respective districts to ensure consistency in data collection, while Coffey staff was on call to address technical issues. Disclaimer: The main purpose of the Pay No Bribe Progress Assessme nt is to assess progress in key perception areas and to re - strategize programmatically if required. The Pay No Bribe Progress Assessme nt was not carried out by an independent third party, but by implementing partners and does not claim scientific rigor. The main objective of the assessment is to generate data for internal discussion and action planning. Nonetheless the programme would li ke to make this report available to the general public for information and transparency purposes.

23. 22 PNB outreach The majority of responden ts ( 79% compared to 73 % in August ) state they ha ve listened to or read information about anti - corruption within the last 3 months. This can be attributed to a more active media engagement of the ACC with regular press releases with updates on ongoing cases (not necessarily related to PNB) . The data suggests no significant impact of gender on the likelihood to listening to or reading about anti - corruption issues . Table 17: Demographics and Respondent who had listened to or read information about anti - corruption by PNB outreach level Gender Age Disability Grand total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes Yes 79% 79% 80% 79% 81% 76% 79% 82% No 15% 15% 15% 16% 14% 15% 15% 14% Do not know 5% 5% 5% 4% 4% 8% 5% 4% The majority of respondents (87 %) listen ing to or read ing about anti - corruption issue s have also discussed it with their friends and family. Similarly , 61 % of the people who ha ve not heard or read about anti - corruption issues , ha ve not discussed it with friends or family within the last 3 months. This represents no change from the previous assessment in August . C hart 48 : Respondent who had listened to or read information about anti - corruption within the last 3 months. C hart 49 : Respondent who had listened to information about anti - corruption per district. C hart 50 : Number of respondent who has discussed anti - corruption issues by respondent who had listened to or read information about anti - corruption Have you discussed anti corruption issues with friends and family in the last 3 months?

22. 21 Respondents in PNB Core areas appear to be significantly more inclined to view the 515 line as safe, compared to non - core areas. The perceived safety of reporting mechanism displays d istrict variance ; similar to the August report ; Bo District ha s the highest percentage of respondents who consider it safe to report to the ACC office – while it is low in Western Area. In Western Area, however, many th ink it safe to report via the PNB mobile app. Table 15 : Demographics and perceived safety of methods of reporting to the ACC Gender Age Disability Values Grand Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes 515 90% 90% 90% 90% 91% 88% 90% 89% PNB Apps 46% 46% 46% 48% 48% 39% 46% 42% Website 8% 9% 7% 11% 7% 5% 8% 9% ACC office 14% 15% 14% 17% 13% 14% 14% 13% I have not made a report 3% 3% 2% 2% 3% 4% 2% 3% C hart 45 : Perceived safeness of reporting methods C hart 46 : Perceived safeness of reporting methods by PNB outreach level C hart 47 : Perceived safeness of reporting methods per district

28. 27 Conclusions Th is third and final Progress Assessment confirms m any of the conclusions arising from the two previous a ssessment s , namely : • P ublic perception s of service provider behaviour is improving ; the vast majority of respondents continue to see positive behaviour change across the three MDA s and in all 5 districts. • PNB Core areas generally perform better th an Non - core and Radio only. This suggest s that CSO outreach has had tangible impact. Where citizens have received intensive outreach , knowledge of PNB and Service Charters is stronger, and the y are more aware of what MDAs should or should not be doing. They are better able to identify behaviour change by service providers and more likely to be able to identify actions taken by a particular MDA in response to public reports. • H igher levels of percei ved positive behavioural change are generally noted in the Provinc es, as compared to Western Area. This tendency, however, reduces by each progress assessment, suggesting that Western Area is slowly catching up with the Provinces. I t is remains significant and suggest s that more intensive engagement in the Provinces continues to yield better results tha n MDA engagement with the national headquarters in Freetown. • Free Quality Education has had a significant impact o n public perception s of the Education secto r. This illustrates the potential of major policy change to redefine the relationship between citizen and service provider. • The ACC has developed a stronger public profile , strengthening its institutional credibility. • Demography does not appear to imp act on people’s perception of bribery; neither by gender, age or disability. The se conclusions reflect the multi - layer ed impact of PNB; on awareness, perception s , and behaviour , both of citizens and service providers . They also highlight the constraint s introduced by limited MDA responsiveness ( especially Western Area) to citizen petty corruption reports . S uccesses, particularly in the Provinces, suggest the need for a strong future focus by the ACC on increas ing MDA accountability in Freetown, continued regular support to district and national IMCs and to the RTWGs, and in particular , the value of locally based agents of accountability, which for PNB have been the CSO Animators.

20. 19 Perception of method of reporting to the ACC 95 % of respondents ha ve knowledge of the 515 hotline ; 44 % of PNB mobile app ; 15 % of the ACC/PNB website ; 37 % that it is possible to repo rt corruption at the ACC office – increas es from the August assessmen t record ing 91%, 30%, 7% and 34% respectively . V ariance between district s remains small , with Western Area ( Rural+Urban ) having the highest frequency of knowledge of the PNB app . R espondents’ interviewed in Urban , Peri - urban and Rural communities expressed similar levels of knowledge on reporting methods, while overall, more respondents from PNB core outreach areas ha ve knowledge of reporting methods than respondents from Hub and Sp okes and Radio Only areas . Data suggests little variance in knowledge of reporting method between gender, ag e group and disability . The only identifiable variance of note is that fewer respondents above the age of 50 hav e knowledge of reporting methods. Ta ble 13 : Demographics and knowledge of methods of reporting to the ACC Gender Age Disability Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes 515 95% 95% 95% 96% 96% 91% 95% 95% PNB apps 44% 44% 44% 47% 44% 40% 44% 41% Website 15% 15% 15% 16% 15% 13% 15% 16% ACC Office 37% 36% 37% 38% 35% 37% 37% 35% C hart 39 : Knowledge of reporting methods to the ACC C hart 40 : Knowledge of reporting methods to the ACC per district C hart 41 : Knowledge of reporting methods to the ACC by Community type C hart 42 : Knowledge of reporting methods to the ACC by PNB outreach level

3. 2 Background The Anti - Corruption Commission of Sierra Leone and Coffey International is, with funding support from UK Aid, implementing the Pay No Bribe Campaign. The Campaign started in September 2016 with the launch of the Pay No Bribe reporting platform. The Pay No Bribe platform is a reporting mechanism for citizens to anonymously report incidents of petty corruption and bribery through a free hotline phone number, mobile apps or the P NB website ( www.pnb.gov.sl 2 ). The platform collects data on bribery and corruption in key service sectors, currently Education, Electricity, Health, Police, Water and the Justice sector. It provides a database on pet ty corruption and bribery trends to support the work of the ACC and MDAs to eliminate corrupt practice. A key PNB strategy has been sustained communication and outreach work in the 5 core districts of Kenema, Kono, Bo, Bombali and Western Area. This is und ertaken both by ACC public education officers and until the end of November 2018 by 123 CSO animators. In addition, public information on petty corruption issues is distributed through IEC and regular radio programming across the country. Purpose The purp os e of the Progress Assessment has been to assess whether the public has experienced a change in the behaviour of public officials with respect to petty corruption including bribery ; whether citizens are now more likely to resist corruption; and whether th e public considers that the ACC is now more likely to act against corruption compared to one year ago. The progress assessment pr ovides a snapshot that has allow ed the Pay No Bribe campaign to assess progress in these key perception areas and to re - strate gize as required. This is the third of the 3 Progress Assessments planned and conducted during 2018. In addition to exploring public perc eption s of behaviour change, the second and third survey s also examine the impact of PNB outreach, perceptions of methods of reporting to the ACC ; and Knowledge of Fees and Services , including source s of information. Methodology The Progress As sessment was undertaken on the 21 th /22 nd /23 rd November 2018 across the 5 core districts - Bo, Bombali, Kenema, Kono and Western Area (Urban and Rural). Within each of these 5 districts the survey focused on: Core Areas receiving regular community outreach; Non - core areas receiving periodic hub and spokes outreach; and areas in receipt of radio outreach only. The questionnai re cover ed the same themes as previous assessment s . The questionnaire was administered by 75 enumerators ( 15 in each of the 5 districts) using Kobo Collect . CSO coordinators receive d an induction to the process and held refresher meeting s with the enumerators on methodology and process. This support ed training received before p revious assessment s . Each enumerator was designated to specific area s identified in the sampling process and requested to collect the following : - 15 responses in a “Core - PNB area” on D ay 1 - 10 responses in a “Limited outreach area” on D ay 2 - 5 responses in an “Area only reached by Radio” on Day 3. 1 The PNB website is scheduled to integrate with the Anti - Corrupti on Commission website ( www.anticorruption.gov.sl ) and will cease to exist as an individual website.

27. 26 PNB animator told me 31% 32% 31% 32% 31% 33% 31% 29% During PNB meeting 19% 18% 19% 18% 18% 19% 18% 16% Other 2% 2% 2% 2% 3% 3% 2% 4% Source of information on fees and services for police Service Charters remain an uncommon source of information for police fees and services . Just 7% of total respondents , and 23% of those claiming to have information on police fees and services accessed that information from a Service Charter. PNB Animators and Radio remain the most c ommon source of information. The demographics d o not indicate any significant impact on source of information for knowledge of police services. Table 22 : Demographics of total respondent population and the source of the information Grand Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes Displayed at the police station or post (Service Charter) 7% 7% 7% 7% 8% 6% 7% 8% The service provider told me 9% 9% 8% 8% 10% 8% 9% 8% Radio programmes 21% 22% 21% 22% 21% 21% 21% 18% PNB animator told me 23% 23% 24% 24% 22% 25% 22% 23% During PNB meeting 16% 15% 16% 15% 16% 18% 15% 16% Other 2% 3% 2% 2% 2% 3% 2% 4% C hart 58 : Number of respondents indicating the source of information for fees and services for Police

12. 11 Table 6: Type of Community and perceived behavioural change among service providers Health Education Police Rural Peri - Urban Urban Rural Peri - Urban Urban Rural Peri - Urban Urban No 20% 20% 18% 11% 7% 5% 32% 42% 39% Yes 80% 80% 82% 89% 93% 95% 68% 58% 61% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% 100% The only variable that consistently appears to influence perception , is whether citizens reside in a PNB core area, a Hub and Spoke area , or a Radio only area. Throughout, respondent perceived most change in service provider behaviour in PNB Core areas and least in areas only reach by radio. Table 7: Focus of PNB outreach and perceived behaviou r change among service providers Health Education Police PNB Core areas Hub and Spoke Radio only PNB Core areas Hub and Spoke Radio only PNB Core areas Hub and Spoke Radio only No 17% 21% 29% 7% 11% 10% 36% 43% 46% Yes 83% 79% 71% 93% 89% 90% 64% 57% 54% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% 100% District and MDA Comparison Earlier programme assessment s suggest that respondents see the most positive behaviour change from provincial service providers and officials; less so in Western Area . This remains the case, although the gap narrowed between the assessment s . For Health, the gap persist s although performance in the Provinces is only marginally higher than Western Area . In Education, the data from across Western Area indicates a significant improvement ; 62% stat e that officials ha ve stopped taking bribes, compared to 48% in the Provinces (analysis provided on page 8). Public perception of behaviour change by SLP continue s to be more positive in the provinces compared to Western Area. C hart 19 : Perceived behaviour change by health officials in the Provinces C hart 20 : Perceived behaviour change by health officials in Western Area

21. 20 Usefulness of reporting mechanisms 97 . 5 % of respondents conside r one of more of the Reporting mechanisms to be useful with just 2 . 5 % of respondents report ing that n o ne of the reporting method s are useful. T he 515 hotline is overwhelmingly considered to be the most useful reporting mechanism ( 87% ) while only 3% view th e ACC Website as a useful means of reporting. This is worth considering by the ACC when determining which reporting methods to promote during outreach and public education. R esponses with respect to the usefulness of reporting mechanism s correlat es to the intensity of outreac h in the area of the interview, with the highest perceived usefulness in PNB Core Areas, a slightly lower response in Hub and Spoke areas and least in Radio only areas – particularly for the 515 hotline. In line with other indicators; gender, age and disability status ha ve limited relevance to perceived usefulness of reporting method . Only the PNB app indicates any difference b y age categor y , with young er people considering it most useful , as may be expected . Table 14 : Demographics and perceived usefulness of methods of reporting to the ACC Gender Age Disability Grand Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes 515 87% 86% 87% 87% 86% 86% 86% 87% PNB apps 34% 33% 35% 36% 34% 31% 34% 31% Website 3% 3% 4% 3% 4% 3% 3% 2% Office 21% 21% 22% 21% 22% 20% 21% 20% None 3% 3% 2% 3% 3% 2% 2% 5% Safety of reporting mechanisms 9 2% of all respondents conside r the 515 line s afe to use, a significant improvement to 80% recorded in the August assessment. In response to previous reports indicating significant number s do not consider the 515 line to be safe, PNB outreach ha s since focuse d on caller a nonymity in order to reassure the public. T he percentage of respondents who consider it safe to report corruption has since increased , reflect ing improved public perception of and confidence in the ACC . C hart 43 : Perceived usefulness of reporting methods C hart 44 : Perceived usefulness of reporting methods by PNB outreach level

26. 25 In this respect, it is worth noti ng that PNB Core areas ( where MDA facilities are held more accountable by CSO animators) , ha ve a significantly higher rate (47%) of service charter d i splay than other areas. In Hub and Spoke and Radio O nly areas, the Radio was the dominant source of information. Table 2 0 : Demographics of total r espondent population and the source of the information Grand Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes Displayed at the school (Service Charter) 24% 23% 26% 24% 27% 19% 23% 31% The service provider told me 20% 21% 19% 19% 22% 17% 21% 19% Radio programmes 36% 37% 35% 38% 36% 33% 37% 31% PNB animator told me 38% 38% 37% 40% 37% 35% 38% 35% During PNB meeting 21% 21% 20% 22% 20% 22% 21% 21% Other 2% 3% 2% 2% 3% 1% 2% 3% Source of information on fees and services in health The pattern for source of information on fees and services within the health sector i s similar to that of Education. The citing of Service Charters as a source of information increased from 33% in August to 50% in November for people with knowledge of fees and Services. This indicates that services charters are a more common information source for Health providers tha n for the other sectors. PNB has heavily promot ed local services charters in clinics and PHU s, contribut ing t o th is increase. Demographic variances were insignificant. Table 21 : Demographics of total respondent population and the source of the information Grand Total Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes Displayed at the PHU, clinic or hospital (Service Charter) 21% 21% 21% 21% 23% 17% 21% 22% The service provider told me 14% 15% 13% 15% 15% 12% 14% 14% Radio programmes 26% 26% 26% 26% 26% 27% 26% 22% C hart 55 : Number of respondents indicating the source of information for fees and services within Education C hart 57 : Number of respondents indicating the source of information for fees and services within Health C hart 56 : Respondents with knowledge indicating the source of information for fees and services within Education by outreach level

15. 14 Respondent demographics do not indicate significant variance (+/ - 5 %) by age, gender an d d isability to bribery resistance. The main factor appears to the reliance of the service, with +50 years and PWDs to be less likely to resist bribery for health services and youth less likely for education. Table 9 : Demographics and increased likelihood of saying NO to bribery for Health service Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes Not more likely 11% 11% 11% 11% 9% 13% 10% 14% More likely 88% 88% 88% 88% 90% 86% 89% 85% (blank) 1% 1% 1% 1% 1% 1% 1% 1% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% Table 10 : Demographics and increased likelihood of saying NO to bribery for Education Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes Not more likely 8% 7% 8% 10% 6% 7% 8% 8% More likely 91% 91% 91% 89% 93% 91% 91% 92% (blank) 1% 1% 1% 1% 1% 2% 1% 1% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% Table 11 : Demographics and increased likelihood of saying NO to bribery to Police Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes 17% 17% 17% 19% 16% 14% 16% 19% 17% 82% 82% 82% 80% 82% 84% 83% 80% 82% 1% 1% 1% 1% 1% 2% 1% 1% 1% Grand Total 100% 100% 100% 100% 100% 100% 100% 100%

11. 10 There is little variance between Urban, Peri - Urban or Rural Communities in terms of perceived change to police behaviour. R esponses varied between district s , but without discernible pattern or relation ship to other variables. Bombali recorded the high est percentage in cessation of polic e bribe requests (38 %) , with Kene ma lowest (7%). Kenema scored high on other les s positive indicators. Western Area continue s to report little change , with 58 % of respondents stating that No Change ha s occurred (up from 56%) , with just 8% of respondents report ing cessation of bribe taking by the police. General Overall r espondent demographics (gender, age, PWDs, Community type) do not indicate significant difference in the perception of behaviour change, with 2 exceptions: 1. Fewer respondents over the age of 50 and with disability perceived positive change to the behaviour of Health officials (74 - 76%), compared to younger respondents (80 - 83%). 2. More respondents in Urban and Peri - Urban areas perceive positive changes by Education officials (95%), compared t o respondents from Rural areas (89%). Table 3: Demographics and perceived behavioural change among health officials Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes No 19% 17% 20% 17% 18% 24% 17% 26% Yes 81% 83% 80% 83% 82% 76% 83% 74% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% Table 4 : Demographics and perceived behavioural change among education officials Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes No 8% 7% 8% 7% 7% 10% 7% 9% Yes 92% 93% 92% 93% 93% 90% 93% 91% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% Table 5 : Demographics and perceived behavioural change among police officials Total Gender Age Disability Female Male Under 29 Years 30 - 49 Year Above 50 years old No Yes No 38% 40% 36% 38% 39% 37% 38% 41% Yes 62% 60% 64% 62% 61% 63% 62% 59% Grand Total 100% 100% 100% 100% 100% 100% 100% 100% C hart 18 : Perceived behaviour change by police officials in Radio only areas